Asset AllocAtor BDo stoY HAYWArD INVestMeNt MANAGeMeNt
Perfecting the
in-house model
BDO Stoy Hayward Investment Management runs
a series of funds of funds portfolios for its private
clients. Shaun Port explains its asset allocation
process, as well as why attractive valuations do
not mean he is fully invested just yet
By shaun port, chief
investment officer,
Bdo stoy hayward
investment management
We use an in-house asset
allocation model, looking
back as far as data will
allow on a consistent basis.
For us this is data going
back to 1990 across a
range of assets including
commodities, commercial
property and funds of
hedge funds (FOHFs), as
well as the traditional
asset classes, by geography
and credit.
There is always criticism
about how much data you
use. But using hundreds
of years of data does not
necessarily give you the
perfect answer. For example,
with bonds, there have
been periods of around 20
years when their return
over cash was negative,
which is unrealistic.
We include private
equity in the model but not
in any of the allocations,
because we have found
what is available to investors
is no better than a
European small-cap fund.
There are private equity
firms sitting on cash as well
plenty of companies funding
draw-downs by selling
valuable public equity
assets to honour their private
equity commitments.
We model up to 15 asset
classes, not simply bonds
versus equities, to identify
extreme events on a monthby-month
basis. We place
particular emphasis on
ensuring we have accounted
for the under-reporting
of volatility across all the
portfolIo coMpArIsoN
In terms of moving from a balanced to an aggressive portfolio,
we add a higher equity weight. This would also be reflected in the
composition of some of the funds themselves, so we would generally
have a lower weight to some of the more cautious asset classes
– such as bonds and property – and a higher weight to hedge
funds and equities.
Within the equity holdings we would have a greater exposure to
emerging than developed markets in the more aggressive portfolios.
In a balanced portfolio we would not have emerging markets as
a standard allocation, although we may do in the future when we
become more positive about equities in general.
For the more aggressive portfolio you are less constrained on
risk; with a more balanced one you are focused almost entirely on
risk. With an aggressive portfolio you can move up the return scale,
although it is probably not as efficient on a risk perspective – the
aim is for returns so your constraint on risk is much less.
32 PORTFOLIO ADVISER [www.portfolio-adviser.com] DEcEmbER 2008
asset classes; right now this
is particularly important for
property, hedge funds and
even corporate bonds.
Once we have a dataset
we are happy with – that is,
not understating risk anywhere
– we use ‘boot-strapping’,
which involves thousands
of simulations of different
portfolios and return
profiles to test their robustness
when combined.
l Portfolio design
When designing portfolios,
we are looking for excess
gross returns over cash,
including constraints that,
for example, will not let
the model run away with
property and hedge funds.
These less liquid assets
have the best risk-adjusted
returns over the longer
period, even if accounting
for lower risk.
Once we have our
model allocations we stress
test them, to detect scenarios
when the model will
not perform.
We use this as our neutral,
long-term allocation
and add colour where we
think we should be allocating
money. This informs
our long-term, five-year
view, but we add colour to
that based on our macro
view of where we think the
markets are going, what is
cheap, what is expensive.
For example, although on a
long-term basis our portfolio
would suggest we
should have money in commercial
property, we are
not committing money
there right now.
l Commodities
In 2005, we decided all
clients should have access
to commodities, as these
improve the efficiency
of the portfolio. So we
designed our own fund.
Over the long-run the correlation
of commodities to
equities is virtually non-